Input Design Variables Research Articles

Geometric optimisation of magnetorheological valve using feedforward neural networks for distribution of magnetic flux density inside the valve

Magnetorheological (MR) fluids are very popular among different smart materials. MR valves are used in dampers for different engineering applications where magnetostatic analysis plays a vital role in analysing the performance indices of the MR valve. In the present study, single hidden layered feedforward neural network (FNN) model is proposed to predict the magnetic flux density in the MR valve using magnetic field data generated from a finite element (FE) model in ANSYS-APDL software package. Application of neural networks in magnetostatic analysis has not been reported till now. The proposed FNN has two geometric parameters and applied current as its features (input design variables). The number of activation units in the hidden layer of FNN has been determined through error analysis. After the architecture of the FNN is defined, the FNN is trained to get optimal weights for the features and output of the hidden layer. Predictions made by the FNN and FE models are then evaluated using mean squared error method and it is shown that the FNN model’s results agree with FE result fairly well. The fitted FNN model is then utilised in geometric optimisation of MR valve using a combination of genetic algorithm and sequential quadratic programming method for different application-specific weight factors. Optimal results are then compared to the unoptimized one for improvements in the objective function according to the weight factors allotted.

Economic Analysis for Energy Efficient Reactive Distillation

Complex chemical reactions and downstream processing can be performed in reactive distillation column to overcome with equilibrium limitations and to make it an economical process. An existing reactive distillation unit is costly to modify due its complex configuration and existing limitations of structure, space area, etc. Modifying an existing plant is a tedious task and more complex than a new process. Thus, to modify the existing operating conditions and input design variables it is necessary to verify by applying the same in real plant condition. This is done using revamping based on rigorous simulation and optimization in Aspen Plus process simulator. The main form of energy generator used in a distillation column is reboiler which directly affects the utilities such as cooling water, electricity and steam. Therefore, optimizing this reboiler duty to reduce energy losses is done using optimization model analysis tool of Aspen Plus. This reduction of energy demands diminishes the operating cost as per the reduction in utility cost. Now this reduction in operating cost is evaluated corresponding to the optimized reflux ratio and number of plates as obtained using sensitivity analysis tool of Aspen Plus.

Insights from Mathematical Modelling into Energy Requirement and Process Design of Continuous and Batch Stirred Tank Aerobic Bioreactors

Bioreaction kinetics, oxygen transfer and energy modelling were applied to stirred tank aerobic bioreactors. This was done to investigate how key input design variables influence bioreactor size, feed and wasted substrate, and electrical energy requirements for aeration and cooling, and to compare batch and continuous modes of operation. Oxygen concentration in the liquid is a key input design variable, but its selection is challenging as it can result in design trade-offs. Reducing its value caused a decrease in electrical energy requirement, however this tended to increase the working volume of the bioreactor. The minimum or near-to-minimum total energy requirement for oxygen transfer occurred when operating at the onset of flooding throughout the bioreaction time. For typical KS values, continuous mode of operation required a much smaller bioreactor volume, due to higher operating cell concentration, and this is a major advantage of continuous over batch.

Predicting the Required Volume Need for Architectural Work Using Artificial Neural Networks in Hospital Buildings

Decision-making in construction design has an important role. The need for estimation tools of planning and project management aspects needs to develop. This paper discussed the benefits of artificial neural network methodology to overcome the problem of estimated the needs of the volume of wall paired, ceiling worked pairing, and ceramic floor pairing for architectural work at the designed stage of the building. The average architecture cost of state building is 29%-51% of total construction value. Data from 15 projects was used for being trained and tested by Artificial Neural Network (ANN) methods with 5 design input variables. The ANN helped to estimate the value of volume requirement on the architectural working of Pratama Hospital building project in remote areas of Indonesia. Those input variables include building area, average column span distance, the height of the building, the shape of the building, and a number of inpatient rooms. From ANN simulation, the best empirical equation of P2V5 modeling was used to predict the need of hospital architecture work volume at conceptual stage with best ANN structure 5-9-3 (5 input variables, 1 hidden layer with 9 neurons and 3 output) with result of estimation accuracy a maximum of 96.40% was reached.

Stochastic Work Flow Bridges Gap Between Subsurface and Surface Disciplines

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 187462, “Bridging the Gap Between Subsurface and Surface Disciplines—A Tool for the Modern Facilities Engineer,” by Z. Cristea, SPE, Stochastic Asset Management, and T. Cristea, Independent, prepared for the 2017 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, USA, 9–11 October. The paper has not been peer reviewed. This paper presents an unbiased stochastic data-driven work flow in which surface and subsurface uncertainties are accounted for and their effects on facilities design and operational decisions are quantified. Unlike the traditional approach in facilities design where the most-conservative values are used typically as design input variables, the proposed work flow accounts for life-cycle variability and correlations of relevant input data. Introduction Traditional, deterministic facilities-design methodologies, based on deterministic (single-point) conservative conditions and design margins, might not recognize the full spectrum of operational conditions throughout a field’s life cycle, resulting in significant residual risk and waste of resources during operations. Despite efforts to standardize the process of project delivery as much as possible, industry projects remain heavily customized, intermittent (high-variety, low-volume) processes, with a high rate of diversification and complexity. On the other hand, facilities operations are expected to be high-volume, low-variety, ideally continuous processes. In the proposed work flow, deterministic models are established to account for dependencies between design input variables (static variables such as bottomhole pressure and temperature) and the desired objective (static results, such as the chemical-injection rate). However, in field situations, the analyzed variables change because of subsurface and surface events with different levels of uncertainty (e.g., condensate banking, lean gas injection, water breakthrough). Stochastic algorithms are used to create probability distribution functions (PDFs) for all analyzed design input variables (stochastic variables). Stochastic algorithms then are applied on the deterministic model, sampling from the previously defined probability distributions. Stochastic results are assembled into insightful charts and used to analyze the most-relevant variables and their correlations affecting the model objectives. Example Case Equipment and Processes. A deterministic model is created to calculate the baseline methanol-injection rate (QMEOH) required to mitigate the hydrate-formation risk in a wet-gas field. Both the deterministic and subsequent probabilistic modeling work is performed in a system designed for statistical computation and graphics. Design-basis input data (static variables), considered to be conservative and covering all operational scenarios, are used in the deterministic model. Furthermore, the deterministic model assumes chemical injection as a mass/continuous process. In the field, parameters such as bottomhole pressure and flow rates of different phases change over time. These changes are caused by subsurface and surface events that can be predicted with different levels of uncertainty. PDFs are created for all relevant parameters. Similar probability distributions are created for all relevant variables.

Multi-Objective Optimization of a Combined Power Plant Fueled by Syngas Produced in a Downdraft Gasifier

In this paper technical-economic feasibility is analyzed about a gas-steam combined-cycle, fueled by syngas produced in a local downdraft gasifier. At first, the downdraft gasifier model is briefly described, where a biomass mix of Refused-Derived Fuel, RDF, and sewage sludge is transformed into syngas; then, the economic model is introduced with the graph about the economic trends of the most important plant components. Starting from the Design of Experiment (DOE) analysis, this paper proposes a multi-objective optimization methodology to obtain the Pareto Frontier of micro gas turbine technologies. In particular, this study allowed us to determine which micro gas turbine technology is more suitable to attract investment capital. The considered performance parameters are: the nominal power of the cycle; the total plant efficiency; the net revenues; the pay back period. On the other hand, the input design variables are represented by the turbine intake temperature, the compression ratio and the biomass chemical composition.

Experimental investigations, input-output modeling, and optimization of spiking phenomenon in electron beam welding of ETP copper plates

The present study aims to determine the set of optimum process parameters in order to minimize spiking during bead-on-plate welding of ETP copper plates using electron beam. Accelerating voltage, beam current, welding speed, focus distance, amplitude and frequency of oscillations were considered as the input design variables. The spiking was characterized in terms of average penetration and standard deviation of penetration. Statistical regression analysis was conducted to correlate the spiking phenomena with the input parameters of electron beam welding. The reasons behind the spiking phenomenon and the associated porosity defects were also explained. This unconstrained optimization problem was solved using genetic algorithm, particle swarm optimization, and desirability function approach in order to maximize weld penetration and minimize standard deviation of penetration. The obtained optimized results were verified through real experiments and a good agreement between them was achieved. Genetic algorithm was found to perform slightly better than other approaches.

Hull form design optimization of twin-skeg fishing vessel for minimum resistance based on surrogate model

Twin-skeg ship has better hydrodynamic performances than regular ship, however, it is still difficult to obtain an accurate relationship between skeg design and overall hydrodynamic performances. Resistance optimization is the major concern of developing twin-skeg ship. This paper proposes a combined approach for hull form design optimization of twin-skeg ship by using computational fluid dynamics (CFD) calculation and surrogate model. Main design parameters of skeg geometry and arrangement could be determined from the design domain by using the proposed method. Parametric modeling technology is adopted for performing design evaluations in an automatic manner with different design parameter combinations. A twin-skeg fishing vessel is selected as research object. In the proposed method, the sample set for constructing surrogate models is generated by using Optimal Latin Hypercube Sampling (OLHS) method, the corresponding responses are calculated through CFD simulations, and then the surrogate models are constructed by using Kriging modeling method, which represent the mathematical relationship between input design variables (skeg shape design variables) and output objective functions (resistance values under four different working conditions). The functional analysis of variance (ANOVA) is performed to investigate how much influence the design variables have on the objective functions. Finally, a multi-objective evolutionary algorithm (NSGA-II) is used to obtain the optimal solution, which shows 5.4% average decrease in the total resistance than the original design. The CFD calculation results of the optimal solution show that the proposed method can achieve minimum resistance design with high accuracy and low time cost.

A possibilistic finite element method for sparse data

This paper presents a ‘possibilistic’ application of the finite element method (FEM) considering a linear elastic body with sparse information of input design variables. The corresponding uncertainty characteristic is referred to as ‘imprecise probability’. Each input variable is constituted as a possibility distribution comprising a family of probability distributions to account for sparse experimental data. As a special case, we consider isotropic material behaviour with imprecise input variables. Computational approaches involving a ‘probability-possibility’ transformation are applied in order to transform sparse experimental information into corresponding possibility distributions. Finally, in a representative example for a plate with a ring hole the possibility for a quantity of interest is calculated using the alpha-level discretisation technique under consideration of the interaction between shear and bulk modulus.

Eliminate Decision Bias in Facilities Planning

This article, written by JPT Technology Editor Chris Carpenter, contains highlights of paper SPE 187283, “Eliminate Decision Bias in Facilities Planning,” by Z. Cristea, Stochastic Asset Management, and T. Cristea, Consultant, prepared for the 2017 SPE Annual Technical Conference and Exhibition, San Antonio, Texas, USA, 8–11 October. The paper has not been peer reviewed. The complete paper holds the traditional facilities-planning methodologies, heavily based on design-basis documents and biased toward the most-conservative conditions, fail to recognize the entirety of operational conditions throughout the oilfield life cycle, leading to significant residual risk and the wastage of resources in the operations stage. An integrated stochastic approach is proposed, accounting for both subsurface and surface uncertainties and their interrelations throughout field life. Introduction The authors discuss an unbiased, data-driven stochastic work flow addressing the effect of subsurface uncertainties on surface-facilities design and operational decisions. Unlike classical design approaches, in which the most-conservative values are typically used as design input variables and assembled into design-basis documents, the stochastic work flow accounts for design-input-variable distribution and combination throughout the entire system life cycle. An example case is provided in which a flow-assurance risk is managed and chemical consumption optimized in a wet-gas field development. Theory and Definitions Oil and gas engineering projects are typically processes of high variety, low volume, and intermittent productivity, and with a high rate of diversification and complexity. Conversely, oilfield-facilities operations are expected to be continuous, characterized by high volumes and low variety. This expectation is reflected in the approach toward facilities design, where single-point, “conservative” design conditions are proposed and assembled as facilities design-basis documents. This approach frequently fails to recognize the risks and uncertainties associated with oilfield developments. In the proposed work flow, deterministic models are established to account for the dependencies between design input variables {static variables [i.e., bottomhole pressure (BHP) and bottomhole temperature (BHT)]} and the desired objective [static results (i.e., chemical- injection rate)]. In the provided example, the analyzed variables change because of subsurface and surface events with different levels of uncertainty (i.e., condensate banking, lean-gas injection, water breakthrough). Stochastic algorithms are used to create probability-distribution functions (PDFs) for all analyzed design input variables (stochastic variables). Stochastic algorithms are then applied in the deterministic model, sampling from the previously defined probability distributions. Stochastic results are assembled into insightful charts and used to analyze the most-relevant variables and correlations affecting the objective function.

The influence of geometric design variables on the kinematic performance of a surface-guided total knee replacement.

Tibiofemoral geometries in a total knee replacement (TKR) affect the performance of an implant during activities of daily living. The specially shaped components of a surface-guided TKR aim to control the tibiofemoral motion, such that a normal pattern of motion is achieved, even at high flexion angles. The purpose of this study was to assess the influence of the design parameters on the kinematic behavior of such an implant. A combination of design variables was determined that resulted in the least deviation from the design kinematic target. Six major design variables were considered to generate customized surface-guided TKR candidates. The contribution of these variables was evaluated by principal component analysis considering the input design variables and the results of the kinematic performance from a virtual simulation of deep squatting. The tibial internal-external rotation and the anterior-posterior translation of the medial and lateral femoral condyles were recorded for each design candidate. A quantified objective function of the kinematic behavior was used to define the design with a maximum agreement with the target pattern of motion. The location and orientation of the flexion-extension axis and the tibial slope were the most contributing parameters on the modes of variation. On the other hand, the conformity between the lateral guiding arcs had the least contribution. Virtual simulation showed that the current TKR reached deep flexion angles under squat load, while the tibia pivoted around the medial center. The tibial rotation was within the expected range of the IE rotation from healthy joints.

Inclusion of design variability in flexible highway pavement life-cycle cost analysis

This research presents a method that accounts for variability in the principal design input variables for fully flexible highway pavements and assesses their effect on pavement performance. Variability is described by statistical terms such as mean and standard deviation and by its probability density distribution. Statistical characterisation of the variation of asphalt layer thickness, asphalt stiffness and subgrade stiffness is addressed. A model is then proposed that represents an improvement on the method of equivalent thickness for the rapid and repeated calculation of performance life. A Monte Carlo analysis is used to estimate pavement performance life to account for uncertainty of input variables and to calculate the probability of failure of a pavement structure. The output is a statistical assessment of the estimated pavement performance. Rather than the single deterministic result that would be derived by considering average values of input variables, a range of values and probabilities is found for any particular outcome. The probabilistic approach offers a way of incorporating risk assessment considerations that are vital for whole-life-cycle economic analysis and decisions. The paper investigates how variability affects the life-cycle cost of a pavement over a 60-year analysis period.

원심압축기 공력성능 해석 프레임워크 구축

A framework for the aerodynamic performance analysis of a centrifugal compressor was constructed using the nondimensional performance. Because of the much time and efforts for the new design of centrifugal compressors, it is widely used in the industry to assess the performance of the compressor under the new operating conditions based on the existing performance results and to apply some design changes to the existing compressors. When checking the performance of centrifugal compressors for different operating conditions by using the existing results, it is easy to utilize the non-dimensional performance head coefficient. If the framework of the non-dimensional performance coefficient is utilized, it is convenient to input design variables and performance curve data. Also it is easy to perform the non-dimensional analysis and performance review for off-design points. The accuracy of the framework was verified by comparison with the results of two 1-stage centrifugal compressors at off-design conditions.

An active learning radial basis function modeling method based on self-organization maps for simulation-based design problems

The radial basis function has been widely used in constructing metamodels as response surfaces. Yet, it often faces the challenge of accuracy if a sequential sampling strategy is used to insert samples sequentially and refine the models, especially under the constraint of computational resources. In this paper, a sensitive region pursuing based active learning radial basis function (SRP-ALRBF) metamodeling approach is proposed to sequentially exploit the already-acquired knowledge in the modeling process for obtaining a desirable estimation of the relationship between the input design variables and the output response. In this method, the leave-one-out (LOO) errors of each sample point are taken to identify the boundaries of sensitive regions. According to the obtained LOO information, the original design space is divided into some subspaces by adopting the self-organization maps (SOMs). The boundary of the most sensitive region, where the output response is multi-modal or non-smooth with abrupt changes, is determined by the topological graph generated by cluster analysis in SOMs. In the most sensitive region, infill sample point searching is performed based on an optimization formulation. Ten numerical examples are used to compare the proposed SRP-ALRBF with four existing active learning RBF metamodeling approaches. Results show the advantage of the proposed SRP-ALRBF approach in both prediction accuracy and robustness. It is also applied to three engineering cases to illustrate its ability to support complex engineering design.

H2-selective mixed matrix membranes modeling using ANFIS, PSO-ANFIS, GA-ANFIS

The novel contribution of the current study is to employ adaptive neuro-fuzzy inference system (ANFIS) for evaluation of H2-selective mixed matrix membranes (MMMs) performance in various operational conditions. Initially, MMMs were prepared by incorporating zeolite 4A nanoparticles into polydimethylsiloxane (PDMS) and applied in gas permeation measurement. The gas permeability of CH4, CO2, C3H8 and H2 was used for ANFIS modeling. In this manner, the H2/gas selectivity as the output of the model was modeled to the variations of feed pressure, nanofiller contents and the kind of gas, which were defined as input (design) variables. The proposed method is based on the improvement of ANFIS with genetic algorithm (GA) and particle swarm optimization (PSO). The PSO and GA were applied to improve the ANFIS performance. To determine the efficiency of PSO-ANFIS, GA-ANFIS and ANFIS models, a statistical analysis was performed. The results revealed that the PSO-ANFIS model yields better prediction in comparison to two other methods so that root mean square error (RMSE) and coefficient of determination (R2) were obtained as 0.0135 and 0.9938, respectively. The RMSE and R2 values for GA-ANFIS were 0.0320 and 0.9653, respectively, and for ANFIS model were 0.0256 and 0.9787, respectively.

Comprehensive heat transfer correlation for water/ethylene glycol-based graphene (nitrogen-doped graphene) nanofluids derived by artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS)

Herein, artificial neural network and adaptive neuro-fuzzy inference system are employed for modeling the effects of important parameters on heat transfer and fluid flow characteristics of a car radiator and followed by comparing with those of the experimental results for testing data. To this end, two novel nanofluids (water/ethylene glycol-based graphene and nitrogen-doped graphene nanofluids) were experimentally synthesized. Then, Nusselt number was modeled with respect to the variation of inlet temperature, Reynolds number, Prandtl number and concentration, which were defined as the input (design) variables. To reach reliable results, we divided these data into train and test sections to accomplish modeling. Artificial networks were instructed by a major part of experimental data. The other part of primary data which had been considered for testing the appropriateness of the models was entered into artificial network models. Finally, predictad results were compared to the experimental data to evaluate validity. Confronted with high-level of validity confirmed that the proposed modeling procedure by BPNN with one hidden layer and five neurons is efficient and it can be expanded for all water/ethylene glycol-based carbon nanostructures nanofluids. Finally, we expanded our data collection from model and could present a fundamental correlation for calculating Nusselt number of the water/ethylene glycol-based nanofluids including graphene or nitrogen-doped graphene.

Application of mathematical modelling for investigating oxygen transfer energy requirement and process design of an aerobic continuous stirred tank fermenter

Fermentation kinetic and oxygen transfer modelling coupled with energy analysis was applied to investigate how key input design variables influenced fermenter size, feed substrate requirement, wasted substrate and aeration system electrical energy requirement. The study showed that trade-offs and compromises are required to select the values of key input variables that can produce superior process designs in terms of the output variables. For example, reducing steady-state oxygen concentration reduced aeration system energy requirements and associated carbon footprint but increased fermenter size and associated cost. Mathematical modelling can assist in more precisely zoning in quantitatively on the selection of design input variable values that can produce a best compromise between conflicting design output variables. Mathematical modelling can also highlight design sensitivities. For example, if the steady-state sugar concentration is reduced below a certain value, then this can lead to an exponential increase in fermenter volume and associated cost, thus it is prudent to operate on the conservative side of this value.

Development and experimental study of a small-scale compressed air radial inflow turbine for distributed power generation

With ever increasing demand on energy, disturbed power generation utilizing efficient technologies such as compressed air energy storage (CAES) and organic Rankine cycle (ORC) are receiving growing attention. Expander for such systems is a key component and its performance has substantial effects on overall system efficiency. This study addresses such component by proposing an effective and comprehensive methodology for developing a small-scale radial inflow turbine (RIT). The methodology consists of 1-D modelling, 3-D aerodynamic investigation and structural analysis, manufacturing with pioneering technique and experimental testing for validation. The proposed 1-D modelling was very effective in determining the primary geometry and performance of turbine based on parametric studies of turbine input design variables. However with CFD analysis, it was shown that more efficient turbine geometry can be achieved that not only provides more realistic turbine performance by capturing the 3-D fluid flow behaviour but also improves turbine efficiency with the aid of parametric studies of turbine geometry parameters. Turbine efficiency was improved from 81.3% obtained from 1-D modelling to 84.5% obtained by CFD. Accuracy of the CFD model was assessed by conducting experiments on the RIT manufactured with stereolithography technique. The CFD model can predict turbine efficiency and power with accuracy of ±16% and ±13% respectively for a wide range of tested operating conditions. Such results highlights the effectiveness of the proposed methodology and the CFD model can be used as benchmarking model for analyses of small-scale RITs. Besides, it was shown that for such applications, the novel manufacturing technique and employed material are very effective for producing prototypes that assist design decisions and validation of CFD model with reasonable accuracy at reasonable cost and in timely manner.

Analyzing sensitivity measures using moment-matching technique

Sensitivity indices are used to rank the importance of input design variables or components by estimating the degree of uncertainty of output variable influenced by the uncertainty generated from input variables or components. With the advent of highly complex engineering simulation models that describe the relationship between input variables and output response, the need for an efficient and effective sensitivity analysis is more demanding. Traditional importance measures either requires extensive random number generations or unable to measure variables interaction effects. In this article, a generalized approach that can provide efficient and accurate global sensitivity indices is developed. The approach consists of two steps; running an orthogonal array based experiment using moment-matched levels of the input variables followed by a variance contribution analysis. The benefits of the approach are demonstrated through different real life examples.

Variability in Pavement Design

Abstract This paper presents the results of a review on variability of key pavement design input variables (asphalt modulus and thickness, subgrade modulus) and assesses effects on pavement performance (fatigue and deformation life). Variability is described by statistical terms such as mean and standard deviation and by its probability density distribution. The subject of reliability in pavement design has pushed many highway organisations around the world to review their design methodologies, mainly empirical, to move towards mechanistic-empirical analysis and design which provide the tools for the designer to evaluate the effect of variations in materials on pavement performance. This research has reinforced this need for understanding how the variability of design parameters affects the pavement performance. This study has only considered flexible pavements. The sites considered for the analysis, all in the UK (including Northern Ireland), were mainly motorways or major trunk roads. Pavement survey data analysed were for Lane 1, the most heavily trafficked lane. Sections 1km long were considered wherever possible. Statistical characterisation of the variation of layer thickness, asphalt stiffness and subgrade stiffness is addressed. A sensitivity analysis is then carried out to assess which parameter(s) have the greater influence on the pavement life. The research shows that, combining the effect of all the parameters considered, the maximum range of 15th and 85th percentiles (as percentages of the mean) was found to be 64% to 558% for the fatigue life and 94% to 808% for the deformation life.